Hyper-spectral Characteristics of Major Types of Soils in Red Soil Region of Jiangxi Province,China
Author:
Affiliation:

Clc Number:

Fund Project:

the National Natural Science Foundation of China(No.41361049), The Project of State Key Laboratory of Soil and Sustainable Agricultural(No.0812201202) and the Natural Science Foundation of Jiangxi Province(No.20122BAB204012)

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
  • |
  • Comments
    Abstract:

    【Objective】Soil spectrum, as a comprehensive reflection of soil physical and chemical properties, is of important significance to soil quality management, digital soil mapping, analysis of soil properties and soil classification in the red soil regions. 【Method】 In this paper, a total of 443 surface soil samples were collected in the typical red soil regions, e.g. Ji’an County, Yujiang County, Xingguo County and Wanli District of Jiang Province, and visible and near infrared reflectance hyper-spectra (350~2500 nm) of the samples were measured with an ASD spectrometer in laboratory. After the treatment of the spectra with the continuum-removal and second order derivative methods, the spectra of the four major soil subgroups and their subordinate soil families in this region were characterized. Then, 19 characteristic variables, such as spectral reflection and absorption of parent materials, iron oxide (goethite and hematite) minerals, organic matter and clay minerals and hyper-spectral reflectance, were cited as indices for Fastclus cluster analysis of the spectra. 【Result】Results show that the soils in the region varied sharply in spectral reflectance from sub-group to subgroup. In terms of spectral characteristic absorption area, in the 620~740 nm spectral bands, the four subgroups exhibited an order of yellow red soil > red soil > brown red soil > weakly red soil, but in terms of the difference between second order derivatives at 420 nm and at 447 nm, they followed an order of yellow red soil > brown red soil > red soil > weakly red soil. The brown red soil was higher in reflectance in the Vis-NIR spectral range, but lower and wider in spectral absorption intensity in the range of 1 900 nm than all the other subgroups. The weakly red soil was the steepest in spectral curve and the highest in reflectance in the near infrared region, showing two relatively strong absorption peaks at 1 400 nm and in 1 900 nm, separately, and a super strong peak at 2 200 nm. In terms of the peak at 2 200 nm, the sub-group of red soil was similar to the sub-group of reddish soil in spectral curve variation tendency, but lower in absorption intensity at 900 nm, 1 400 nm, 1 900 nm and in 2 200 nm and higher in position of the curve. As a result of variation of the duration of flooding, paddy soils of the four sub-groups of red soils varied greatly in spectral characteristics, and they followed an order of gleyed paddy soil > waterlogged paddy soil > submergic paddy soil in terms of spectral reflectance; an order of submergic paddy soil > waterlogged paddy soil > gleyed paddy soil in terms of spectral absorption area in the range of 620 nm~740 nm; and an order of gleyed paddy soil > submergic paddy soil > waterlogged paddy soil in terms of difference between second order derivatives at 420 nm and at 447 nm. Classification, based on the 19 indices, of red soils in the region by soil subgroup reached 86.23% in accuracy and by soil family 66.37%. 【Conclusion】Obviously soil Vis-NIR spectral characteristics can be used as quantitative indices for reference in classification by sub-group of the red soils in Jiangxi Province.

    Reference
    Related
    Cited by
Get Citation

ZHAO Xiaomin, YANG Meihua. Hyper-spectral Characteristics of Major Types of Soils in Red Soil Region of Jiangxi Province, China[J]. Acta Pedologica Sinica,2018,55(1):31-42.

Copy
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Received:November 25,2016
  • Revised:July 27,2017
  • Adopted:September 25,2017
  • Online: October 30,2017
  • Published: